something is wrong with my times.....

[Shadow ÒÓ]

this HAS to be wrong.....

average time per frame 10.13m on a P III 1gig 512 megs of ram
average time per frame 3.41m on a Duron 600 256 megs of ram

According to FoldMonitor

now I'll admit I'm really gaining respect with my first AMD system, and waiting on memory for my new XP 1800......but if AMD is THAT much faster at folding, why didn't one of you guys smack me and wake me up a long time ago?

This is after running folding for around 18 hours.....so the times should be pretty accurate for the AMD...the Intel has been running for days now.

[Allan Nielsen]

Shadow, have you made sure that the proteins they are working on are the same? My P3@1 gig does a Villin frame in 9 minutes with the console text only version. AMD is faster than Intel in folding, but not THAT faster!

I suspect the AMD is crunching on a bba5_build_ext unit, while the P3 is crunching on a NativeVillin.

[Ploaf]

AMD is much faster than Intel at this. I don't know if a Duron 600 is going to be all that much faster than a PIII@1000 though. They should be closely matched at those speeds.
It could be that they were working on two different proteins at the time. Some are smaller than others, but they are all divided into 100 frames. Smaller proteins will just finish a frame more quickly. I'm not sure but I think that the avg time per frame is only for that particular WU and is reset when a new WU is downloaded.
Different operating systems produce different times too. Win98SE is about the fastest for folding. WinXP/2000 usually has more running in the background, as does Linux and they are a little slower.

<Edit>DOH! Allan beat me to it :D</edit>

[Shadow ÒÓ]

ahh....one is working on NativeVillin (P III) and the AMD is working on bba5_build_ext....duh.

Thanks for the quick responses.

[Allan Nielsen]

Originally posted by Ploaf

DOH! Allan beat me to it :D

Hehehehe
:beer: